COGNITIVE NEUROSCIENCE, 2016 Vol. 7, Nos. 1–4, 128–137, http://dx.doi.org/10.1080/17588928.2015.1053441

Effects of contrast inversion on depend on gaze location: Evidence from the component

Katie Fisher, John Towler, and Martin Eimer

Department of Psychological Sciences, Birkbeck College, University of London, London, UK

Face recognition is known to be impaired when the contrast polarity of the eyes is inverted. We studied how contrast affects early perceptual face processing by measuring the face-sensitive N170 component to face images when the contrast of the eyes and of the rest of the face was independently manipulated. Fixation was either located on the eye region or on the lower part of a face. Contrast-reversal of the eyes triggered delayed and enhanced N170 components independently of the contrast of other face parts, and regardless of gaze location. Similar N170 modulations were observed when the rest of a face was contrast-inverted, but only when gaze was directed away from the eyes. Results demonstrate that the contrast of the eyes and of other face parts can both affect , but that the contrast polarity of the eye region has a privileged role during early stages of face processing.

Keywords: Face perception; contrast inversion; N170 component; event-related brain potentials.

Reversing the contrast polarity of familiar and serve as an “anchor” for holistic face processing dramatically impairs their recognizability (e.g., (Nemrodov, Anderson, Preston, & Itier, 2014; Galper, 1970; Johnston, Hill, & Carman, 1992). This Rossion, 2009; see also Orban de Xivry, Ramon, effect appears to be specific to face perception, as the Lefèvre, & Rossion, 2008). In line with this recognition of non-face objects is much less sensitive hypothesis, it has been shown that fixating gaze on to contrast reversal (Nederhouser, Yue, Mangini, & the nasion region (between and just below the eyes) is Biederman, 2007; Vuong, Peissig, Harrison, & Tarr, beneficial for many face processing tasks (Peterson & 2005). Changing contrast polarity may Eckstein, 2012). The first fixation on a face image is Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 disproportionately affect face recognition because it typically directed to this region (Hsiao & Cottrell, removes shading and pigmentation information that is 2008). Inverting the contrast of the eyes should critical for the discrimination of facial identity (e.g., therefore have a greater effect on face processing Kemp, Pike, White, & Musselman, 1996; Liu, Collin, than contrast inversions of other parts of a face, in Burton, & Chaudhuri, 1999; Liu, Collin, & particular when eye gaze is directed toward its Chaudhuri, 2000; Russell, Sinha, Biederman, & preferred position near the eye region. Nederhouser, 2006). The eye region in particular Gilad, Meng, and Sinha (2009) explored this contains contrast-related signals that are relevant for hypothesis with “contrast chimera” faces, which face recognition. For this reason, the eyes may be include both contrast-inverted and contrast-normal prioritized during the structural encoding of faces, regions, and demonstrated that restoring only the

Correspondence should be addressed to: Martin Eimer, Department of Psychological Sciences, Birkbeck College, University of London, Malet Street, London WC1E 7HX, UK. E-mail: [email protected]. Our thanks to Bruno Rossion’s lab for sharing their face images with us. No potential conflict of interest was reported by the authors. This work was supported by the Economic and Social Research Council (ESRC), UK [grant number ES/K002457/1].

© 2015 Taylor & Francis FACE PERCEPTION AND CONTRAST INVERSION 129

eye region of a contrast-inverted face to positive position near the eye region. A recent study contrast improved recognition performance to (Nemrodov et al., 2014) has demonstrated that N170 approximately 90% of the level observed with face-inversion effects are strongly modulated by the contrast-normal faces (see also Sormaz, Andrews, & current location of fixation (see also De Lissa et al., Young, 2013). They also showed that fMRI activity in 2014, for similar findings), and this might also be the the elicited by these positive-eyes case for N170 modulations that are triggered by chimeras was indistinguishable from the response to changing the contrast polarity of faces or face parts. contrast-normal faces. Such observations suggest that In the present experiment, we varied the contrast of the effects of contrast inversion on face processing the eye region and the contrast of the rest of a face might be primarily or even exclusively driven by the orthogonally and also manipulated fixation location, to contrast of the eye region. In the present experiment, test whether N170 components are exclusively sensitive we tested this hypothesis by measuring the face- to eye contrast, and whether this depends on eye gaze sensitive N170 component of the event-related being directed toward the eye region. Participants potential (ERP). The N170 is an enhanced negativity performed a one-back repetition detection task with at lateral occipital-temporal electrodes that emerges contrast-normal faces, fully contrast-inverted faces, and 150–200 ms post-stimulus in response to faces as two types of contrast chimeras where either the contrast compared to non-face objects (e.g., Bentin, Allison, of the eye region or the rest of the face was inverted Puce, Perez, & McCarthy, 1996; Eimer, 2000; see, (negative-eyes and positive-eyes chimeras; see 2011; for review). Because the N170 is generally Figure 1A). This independent manipulation of the unaffected by face familiarity (Eimer, 2000), it is contrast of the eye region and of the rest of the face interpreted as a neural marker of the perceptual allowed us to determine the relative effects of contrast- structural encoding of faces that precedes their inverting either of these regions on N170 amplitudes and recognition. The N170 is highly sensitive to face latencies. One face image was presented at a time, and inversion and to manipulations of face contrast. appeared unpredictably either in the upper or lower Upside-down faces and contrast-inverted faces elicit visual field, so that eye gaze was either centered delayed and enhanced N170 components relative to between both eyes (Upper fixation condition) or upright or contrast-positive faces (e.g., Itier, Latinus, between the nose and mouth (Lower fixation condition; & Taylor, 2006; Itier & Taylor, 2002). Such N170 see Figure 1B). N170 contrast inversion effects were modulations have been attributed to disruptive effects measured separately for both fixation conditions to find on perceptual face processing caused by changing the out whether these effects are modulated by gaze location, typical orientation or the contrast polarity of faces that is, by the relative distance between fixation and the (Itier et al., 2006; Rossion et al., 1999). contrast-inverted part of a face. Several ERP studies have shown that early face processing might be particularly sensitive to the contrast polarity of the eye region. Itier, Alain, METHODS Sedore, and McIntosh (2007) found that N170 Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 modulations elicited by contrast-inverted as Participants compared to contrast-positive faces were eliminated for faces without eyes. Along similar lines, a recent Fourteen participants (10 female) aged 20–39 years ERP study with contrast chimera faces (Gandhi, (mean age 28 years) took part in the study. Their face Suresh, & Sinha, 2012) showed the usual N170 recognition abilities were tested with the Cambridge delay and enhancement for fully contrast-inverted Face Memory Test (Duchaine & Nakayama, 2006). faces, but found that the N170 to positive-eyes All scores were within ±1 standard deviation of the chimeras (where the eye region appeared in normal mean. This study was approved by the ethics contrast) was statistically indistinguishable from the committee of the Department of Psychological N170 component to contrast-normal faces. These Sciences, Birkbeck College, University of London. observations suggest that early perceptual face processing stages that are reflected by the N170 are exclusively sensitive to the contrast polarity of the eye Stimuli and procedure region, and remain unaffected by the contrast of other parts of the face. However, because Itier et al. (2007) Photographs of 25 male and 25 female faces (front and Gandhi et al. (2012) did not manipulate gaze view; neutral expression; external features removed) location, these studies could not determine whether were employed, with permission from Bruno this differential effect depends critically on a fixation Rossion’s lab, where these face images were created 130 FISHER ET AL.

Figure 1. (A) Example of four different face contrast types tested (contrast-normal faces, contrast-inverted faces, positive-eye chimeras, negative-eye chimeras). For the positive-eyes chimeras, the face outside the eye region appeared in negative contrast. For the negative-eyes chimeras, the eye region was contrast-inverted and the rest of the face was contrast-normal. (B) Illustration of the stimulation procedure. Each face was presented for 200 ms, and there was an interval of approximately 1450 ms between two successive face presentations. Faces appeared randomly and unpredictably in a lower or upper position, so that participants’ gaze was either on the upper part of the nose (Upper fixation condition), or on the area between the nose and the mouth (Lower fixation condition). In the example shown, a lower-fixation face is followed by a (non-matching) upper-fixation face.

and first used (Laguesse, Dormal, Biervoye, Kuefner, which has a spatially restricted circular measurement Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 & Rossion, 2012). Four different contrast versions window of approximately 1°. Because the experiment (Figure 1A) were generated for each face, using included two fixation conditions (fixation centered Adobe Photoshop. Contrast-normal images were between both eyes or between nose and mouth), two created by converting the original color images into luminance values within two spatially corresponding grayscale, and adjusting their luminance to a constant measurement windows were obtained for each face level. Image contrast was inverted to produce fully contrast type. Measurement windows were centered contrast-inverted faces. Negative-eyes chimeras were either on the nasion or the philtrum (the central ridge constructed by contrast-inverting a horizontal section between the nose and the mouth) of the faces, across the eye region of contrast-normal faces which respectively. For nasion-centered measurements, included the eyes, lower eye socket, nasion, and average luminance values were 12.30 cd/m2 (contrast- eyebrows. The transition between the contrast- normal faces), 18.12 cd/m2 (fully contrast-inverted normal and inverted regions was smoothed in faces), 17.85 cd/m2 (negative-eyes chimeras), and Photoshop to avoid abrupt contrast polarity changes. 12.28 cd/m2 (positive-eyes chimeras). For philtrum- Negative-eyes chimeras were contrast-inverted to centered measurements, average luminance values produce positive-eyes chimeras. were 10.47 cd/m2 (contrast-normal faces), 21.22 cd/m2 Luminance values for the four different face contrast (fully contrast-inverted faces), 10.47 cd/m2 (negative- types were recorded from a viewing distance of 100 cm eyes chimeras), and 21.17 cd/m2 (positive-eyes with a Konica Minolta CS-100A color/luminance meter, chimeras). Faces were presented against a gray FACE PERCEPTION AND CONTRAST INVERSION 131

background (4.92 cd/m2). The visual angle of all face (blinks and vertical eye movements), and voltages at images was 3.55° × 2.76°. any electrode exceeding ±80 µV (movement artifacts) All face stimuli were shown on a CRT monitor for were removed from analysis. EEG was averaged 200 ms at a viewing distance of 100 cm. The intertrial relative to a 100-ms pre-stimulus baseline for each interval varied randomly between 1400–1500 ms. A combination of stimulus type (contrast-normal, black fixation cross (size: 0.60° × 0.60°) remained on contrast-inverted, positive-eyes chimera, negative- the screen throughout each experimental block. Faces eyes chimera) and fixation position (upper, lower). appeared either in the upper or lower visual field, Only non-target trials (i.e., trials where the randomly intermixed across trials, with a vertical immediately preceding image was not repeated) displacement relative to central fixation of ±1.35° were included in the ERP analyses. N170 peak (Figure 1B). For faces in the upper visual field, the latencies were computed at lateral posterior fixation cross was located on the philtrum (Lower electrodes P9 and P10 (where this component is fixation condition). For faces in the lower visual field, maximal) within a 150–200-ms post-stimulus time fixation was centered on the nasion (Upper fixation window. N170 mean amplitudes were calculated for condition). Participants were instructed to maintain the same electrode pair and time window. Additional gaze on the central fixation cross throughout each analyses were conducted for P1 peak latencies block. The experiment included 10 blocks of 80 (measured within an 80–130-ms post-stimulus time trials, resulting in a total of 800 trials. Each of the window). All t-tests comparing N170 latency or eight combinations of stimulus type (contrast-normal, amplitude differences between stimulus types were contrast-inverted, positive-eyes chimera, negative- Bonferroni-corrected, and corrected p-values are eyes chimera face type; Figure 1A) and stimulus reported. location (upper versus lower; Figure 1B) appeared on 90 randomly distributed trials throughout the experiment. Repetitions of the same face image RESULTS across successive trials were not allowed on these trials. On the remaining 80 randomly interspersed Behavioral performance trials, the image that was presented on the preceding trial was immediately repeated at the same location. Participants detected 81% of all immediate face Participants performed a one-back matching task, and stimulus repetitions in the one-back task. Mean responded with a right- or left-hand button press response time (RT) on these target trials was 618 ms. (counterbalanced across participants) to immediate Therewerenosignificant differences between the four stimulus repetitions. Following the main experiment, face types (normal, inverted, positive-eyes or negative- participants completed the Cambridge Face Memory eyes chimera) for RTs, F(3,39)=2.6,orerrorrates, Task, where the faces of six target individuals shown F(3, 39) = 1.7, on these infrequent target trials. from different viewpoints have to be memorized and then distinguished from distractor faces (see Duchaine Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 & Nakayama, 2006, for a detailed description). ERP components

Upper fixation EEG recording Figure 2 shows ERP waveforms measured at EEG was recorded using a BrainAmps DC amplifier lateral posterior electrodes P9/P10 in response to with a 40 Hz low-pass filter and a sampling rate of faces that appeared in the lower visual field (Upper 500 Hz from 27 Ag-AgCl scalp electrodes. Electrodes fixation condition). The contrast polarity of the eye at the outer canthi of both eyes were used to record region affected N170 latencies and amplitudes, with the horizontal electroculogram (HEOG). During delayed and enhanced N170 components to faces recording, EEG was referenced to an electrode on with contrast-inverted eyes, and this was the case the left earlobe, and was re-referenced offline regardless of whether the contrast of the rest of the relative to the common average of all scalp face was normal or inverted (Figure 2, top panels). electrodes. Electrode impedances were kept below Changing the contrast polarity of the rest of the face 5kΩ. The EEG was epoched from 100 ms before to did not affect N170 amplitudes or latencies when the 250 ms after face stimulus onset. Epochs with HEOG contrast of the eye region was held constant (Figure 2, activity exceeding ±30 µV (horizontal eye bottom panels). In other words, with fixation on the movements), activity at Fpz exceeding ±60 µV nasion, N170 modulations were driven entirely by the 132 FISHER ET AL.

Upper Fixation

Effect of Eye Contrast: Positive Face

–10µV N170

250ms 250ms

P9 P10 8µV Negative Eyes Effect of Eye Contrast: Negative Face Positive Eyes

250ms 250ms

P9 P10

Effect of Face Contrast: Positive Eyes

250ms 250ms

P9 P10

Negative Face Effect of Face Contrast: Negative Eyes Positive Face

250ms 250ms

Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 P9 P10

Figure 2. Grand-averaged ERPs elicited at lateral temporo-occipital electrodes P9 (left hemisphere) and P10 (right hemisphere) in the 250-ms interval after stimulus onset in the Upper fixation condition. Ticks on the time axes represent 50 ms intervals. Top panels: Effects of the contrast of the eye region (negative eyes vs. positive eyes) on N170 components, shown separately for face images where the area outside the eye region was contrast-normal (positive face) or contrast-inverted (negative face). Bottom panels: Effects of the contrast of the rest of the face (negative face vs. positive face) on N170 components, shown separately for face images where the eye region was contrast-normal (positive eyes) or contrast-inverted (negative eyes).

contrast of the eye region, but remained unaffected by contrast (positive, negative), Face contrast (positive, the contrast polarity of the rest of the face. In addition negative), and Hemisphere (left, right). There were to these N170 differences, eye contrast also affected significant effects of Eye contrast on N170 latencies, 2 the peak latency of the earlier P1 component, which F(1, 13) = 108.72, p < .001, ηp = .89, and N170 2 appeared to be delayed specifically for face images amplitudes, F(1, 13) = 21.31, p <.001,ηp =.62, with contrast-inverted eyes (Figure 2, top panels). reflecting delayed and enhanced N170 components These observations were confirmed by analyses for faces with contrast-inverted eyes. These effects of N170 peak latencies and mean amplitudes with did not interact with Hemisphere, both F <2.2. repeated-measures ANOVAs with the factors Eye Analyses of N170 peak latencies (collapsed across FACE PERCEPTION AND CONTRAST INVERSION 133

electrodes P9 and P10) confirmed N170 delays for suggesting that the effects of eye and face contrast face images with negative versus positive eyes both on the latency of the N170 were independent and when the rest of the face was positive (170.4 ms vs. additive. N170 peak latencies (collapsed across P9 163.4 ms; t(13) = 7.7, p < .001) or negative and P10) were reliably delayed for negative versus (170.4 ms vs. 162.6 ms; t(13) = 9.7, p < .001). positive face contrast images (Figure 3,bottom Likewise, N170 amplitude enhancements for faces panel) both when the eyes were positive (164.4 ms with negative versus positive eyes were reliable both vs. 161.9 ms; t(13) = 3.61, p < .006) or negative when the rest of the face was positive (1.55 μV; (168.6 ms vs. 165.6 ms; t(13) = 3.97, p <.004).A t(13) = 4.23, p < .002) or negative (1.07μV; significant interaction between Eye contrast and t(13) = 3.97, p < .004). The absence of interactions Hemisphere on N170 latency, F(1, 13) = 5.26, 2 between eye and face contrast, both F < 1.2, showed p <.05,ηp = .29, was due to the fact that the that the contrast polarity of the rest of the face did N170 delay caused by negative eyes was largely not affect the delay and enhancement of N170 confined to the right hemisphere (see Figure 3,top components to negative eyes. There were also no panel). At right-hemisphere electrode P10, this delay main effects of Face contrast on N170 latency, for negative versus positive eyes was present when F < 1.2, or amplitude, F <3.2,confirming that the the rest of the face was positive (165.4 ms vs. contrast polarity of the rest of the face had no 161.3 ms; t(13) = 4.08, p < .003) or negative impact on N170 components in the Upper fixation (170.3 ms vs. 164.0 ms; t(13) = 4.75, p < .002). condition. There were no corresponding N170 delays over the An analysis of P1 peak latencies with the factors left hemisphere (both t < 1.6). For N170 mean Eye contrast, Face contrast, and Hemisphere revealed amplitudes, there were significant effects of both 2 asignificant effect of Eye contrast, F(1, 13) = 11.53, Eye contrast, F(1, 13) = 19.4, p < .001, ηp =.60, 2 p < .005, ηp =.47,confirming that the P1 component and Face contrast, F(1, 13) = 15.9, p <.001, 2 was delayed for faces with contrast-inverted eyes. ηp = .55, and an interaction between these two 2 There was no main effect of face contrast, and no factors, F(1, 13) = 8.4, p < .05, ηp =.39.N170 interactions between eye or face contrast and amplitude enhancements to negative versus positive hemisphere, all Fs < 1.9. To test whether the N170 eyes (Figure 3, top panel) were present both when delay for faces with contrast-inverted as compared to the rest of the face was positive (1.81 µV; contrast-normal eyes can be completely accounted for t(13) = 4.83, p < .001) and negative (0.91 µV; by the delay of the preceding P1 component to t(13) = 2.90, p < .04). However, the contrast of the contrast-inverted eyes, we performed an additional rest of the face affected N170 amplitude only when analysis of P1-N170 peak-to-peak differences the eyes were positive (1.27 µV; t(13) = 4.33, (obtained by subtracting P1 peak latencies from p < .002), but had no significant differential effect N170 peak latencies) for the factors Eye contrast, for faces with negative eyes (0.37 µV; t <1.8). Face contrast, and Hemisphere. A main effect of P1 peak latencies were not systematically affected 2 Eye contrast, F(1, 13) = 8.21, p <.013,ηp =.39, by the contrast polarity of the eyes or the rest of the Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 confirmed that the N170 delay in response to face (see Figure 3). There were no significant effects contrast-inverted eyes remained reliably present even of Eye contrast, Face contrast, or interactions between when the corresponding earlier P1 delay is taken into these two factors and Hemisphere in the Lower account. fixation condition, all F < 2.9.

Lower fixation Comparison between fixation conditions

Figure 3 shows ERP waveforms measured at P9/ The comparison of Figures 2 and 3 shows that gaze P10 to faces in the upper visual field (Lower fixation location strongly affects how N170 components are condition). In contrast to the Upper fixation modulated by inverting the contrast polarity of the condition, N170 latencies and amplitudes were eye region and the rest of the face. This was systematically affected not only by eye contrast confirmed by additional analyses that included (upper panel), but also by the contrast polarity of Fixation (nasion vs. philtrum) as an additional factor. the rest of the face (lower panel). There were main An interaction between Eye contrast and Fixation for 2 effects of Eye contrast, F(1, 13) = 13.4, p <.001, N170 latency, F(1, 13) = 22.53, p =.001,ηp =.63, 2 ηp = .51, and Face contrast, F(1, 13) = 20.5, was due to the fact that the N170 delay caused by 2 p < .001, ηp = .61, on N170 latency, and no inverting the eye region was twice as large with upper interaction between these factors, F <1.32, fixation (7.3 ms) than lower fixation (3.7 ms). There 134 FISHER ET AL.

Lower Fixation

Effect of Eye Contrast: Positive Face

–10µV N170

250ms 250ms

P9 P10 8µV Negative Eyes Effect of Eye Contrast: Negative Face Positive Eyes

250ms 250ms

P9 P10

Effect of Face Contrast: Positive Eyes

250ms 250ms

P9 P10

Negative Face Effect of Face Contrast: Negative Eyes Positive Face

250ms 250ms Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016

P9 P10

Figure 3. Grand-averaged ERPs elicited at lateral temporo-occipital electrodes P9/10 in the 250-ms post-stimulus interval in the Lower fixation condition. Ticks on the time axes represent 50 ms intervals. Top panels: Effects of eye contrast (negative eyes vs. positive eyes) on N170 components, for face images where the area outside the eye region was contrast-normal (positive face) or contrast-inverted (negative face). Bottom panels: Effects of the contrast of the rest of the face (negative face vs. positive face) on N170 components, for face images where the eye region was contrast-normal (positive eyes) or contrast-inverted (negative eyes).

was also an interaction between Face contrast and eye region enhanced N170 amplitudes regardless of 2 Fixation F(1, 13) = 22.58, p <.001,ηp =.63,as gaze location. However, the effect of Face contrast on inverting the contrast of the rest of the face delayed N170 amplitude was modulated by Fixation 2 the N170 in the Lower fixation condition, but had no F(1, 13) = 6.26, p <.05,ηp = 0.32, as inverting the impact on N170 latency with upper fixation. For N170 contrast of the rest of the face enhanced N170 amplitude, Eye contrast did not interact with Fixation, amplitude with lower fixation, but had no effect in F < 1, demonstrating that inverting the contrast of the the Upper fixation condition. FACE PERCEPTION AND CONTRAST INVERSION 135

DISCUSSION differences between the four face contrast types used in this study (see Liu et al., 1999, for a study where Inverting the contrast polarity of face images impaired the luminance of normal and contrast-inverted faces early stages of perceptual face processing, and this was controlled), but appears to be specific to the was reflected by delayed and enhanced face-sensitive contrast polarity of the eyes when gaze is focused N170 components for contrast-inverted as compared nearby. to contrast-normal faces. Recent studies have shown Inverting the contrast of face parts outside the eye that restoring the normal contrast polarity of the eye region can also modulate perceptual face processing, region while the rest of the face remains contrast- as reflected by the N170 component, but this depends inverted improves recognition performance (Gilad critically on gaze direction. With fixation located et al., 2009; Sormaz et al., 2013) and can eliminate between the nose and mouth, N170 components inversion-related N170 modulations (Gandhi et al., were delayed and enhanced when the rest of the 2012). The present experiment demonstrated that face was contrast-negative. The N170 delay was contrast-inversion of the eyes or of the rest of the independent of the contrast polarity of the eye face can both affect N170 components, depending region, while the N170 amplitude enhancement was on which part of a face is fixated. only reliable for faces with positive eyes (Figure 3, Inverting the contrast polarity of the eye region bottom panel). When gaze was focused on the upper delayed and enhanced the N170. These effects were part of a face between the two eyes, the contrast not modulated by the contrast polarity of the rest of polarity of the rest of the face had no impact on the face, regardless of whether fixation was centered N170 amplitudes and latencies (Figure 2, bottom between the eyes or on the lower part of the face. This panel). The fact that the contrast inversion of face observation that N170 latency and amplitude parts outside the eye region affected N170 modulations triggered by negative eyes are components only in the Lower fixation condition independent of the contrast polarity of other face again shows that the perceptual analysis of faces is parts confirms and extends earlier observations highly sensitive to image contrast near the currently (Gandhi et al., 2012), and demonstrates that fixated location. The contrast polarity of face parts changing the polarity of the eye region affects outside the eye region is not always irrelevant for perceptual face processing regardless of the polarity early stages of perceptual face processing, but will of the rest of the face. Although negative eyes elicited affect face perception when gaze is directed away delayed N170 components in both fixation conditions, from the eyes. The fact that restoring the normal this delay was larger when gaze was centered near the polarity of the eye region in contrast chimera faces eyes (Upper fixation) than when gaze was centered does not improve face recognition up to the level below the nose (Lower fixation condition). In observed with normal faces (Gandhi et al., 2012; addition, the N170 delay for negative eyes was Sormaz et al., 2013) may be linked to the effects of present bilaterally in the Upper fixation condition, inverting the contrast of the rest of the face on but was restricted to the right hemisphere with lower perceptual face processing, as demonstrated in this Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 fixation. This shows that the proximity of the eye experiment during lower fixation. region to current fixation modulates the degree to Although the contrast of the rest of the face can which contrast inversion of this region affects affect perceptual face processing, our results also perceptual face processing. emphasize the central importance of contrast It should be noted that the larger N170 delay for information from the eye region. The fact that N170 negative versus positive eyes in the Upper fixation modulations triggered by contrast-inverted eyes were condition may at least in part be due to the fact that robustly present in both fixation conditions, and the the earlier P1 component was also reliably delayed for observation that they were independent of the contrast faces with negative eyes in this condition, although of the rest of the face, both demonstrate a privileged the N170 delay remained significant even when the status of the eye region that is already apparent during P1 latency difference was taken into account. A P1 early face processing stages between 150 and 200 ms delay in response to contrast-inverted as compared to post-stimulus. In addition to comparing the effects of contrast-normal stimuli has been observed before contrast-inverting the eye region with the effects of (Itier & Taylor, 2002). The fact that this P1 delay inverting the whole of the rest of the face, as was was only found in the Upper fixation condition done in the present study, future research should also when the eye region was contrast-inverted shows investigate how manipulating only the contrast of one that it is not simply a result of the luminance specific face part outside the eye region (such as the 136 FISHER ET AL.

mouth) affects the N170 component, whether there the default setting during face perception (Hsiao & are systematic differences in N170 contrast-inversion Cottrell, 2008), the face-processing system may be effects between facial features, and how this is particularly sensitive to contrast information that affected by gaze direction and current task demands. originates from a region that extends horizontally It should also be noted that no eye tracker was from fixation into the left and right visual field. Such employed in the present study to continuously a retinotopic bias would account for the fact that monitor the precise location of fixation on the contrast-inverting the rest of the face affects N170 nasion or philtrum, and that there may have been components only in the Lower fixation condition, subtle differences in gaze direction between where the inverted areas fall within the critical individual upper or lower fixation trials. However, retinotopic region next to fixation (see also Chan, the fact that the vertical position of each face Kravitz, Truong, Arizpe, & Baker, 2010, for further stimulus was unpredictable rules out anticipatory evidence for retinotopic biases in face processing). It gaze adjustments, and makes it highly likely that will also be important to determine whether any such participants’ gaze was close to the nasion or biases toward particular retinotopic regions in face philtrum on the majority of all trials. processing may be modulated by top-down factors The current results show that structural encoding of such as selective spatial attention. faces is highly sensitive to contrast signals from the Overall, this study has demonstrated the eye region, and that contrast information from other importance of contrast signals for early stages of parts of the face may affect perceptual processing perceptual face processing. Because contrast primarily when eye gaze is directed away from the differences between different face parts remain eyes to lower parts of a face image. Reversing the constant under a wide variety of lighting conditions, contrast polarity of faces generally impairs face they are critical for the rapid detection of a generic perception and recognition, and this has been face template in the visual field (Sinha, 2002). attributed to the disruption of information that can Contrast information from the eyes has a privileged be derived from skin pigmentation (e.g., Vuong role, and focusing gaze and selective attention on the et al., 2005), or of three-dimensional shape-from- eye region therefore provides the optimal reference shading information that is important for point for the construction of contrast-sensitive representing facial shape (e.g., Johnston et al., representations during the perceptual structural 1992). Reversing the contrast polarity of the eye encoding of faces. region is particularly disruptive, because this region contains several contrast-related signals (the Original manuscript received 19 November 2014 boundaries between the sclera, iris, and pupil of the Revised manuscript received 5 May 2015 eye, contrast differences between the eyes and First published online 18 June 2015 surrounding regions, and the shape of the eyebrows) that appear to be critical for face-detection and recognition processes (e.g., Gilad et al., 2009;

Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 REFERENCES Peterson & Eckstein, 2012; Sormaz et al., 2013). The delay and enhancement of the face-sensitive Bentin, S., Allison, T., Puce, A., Perez, E., & McCarthy, G. N170 component to faces with contrast-inverted eyes (1996). Electrophysiological studies of face perception may thus be interpreted as an electrophysiological in . Journal of Cognitive Neuroscience, 8, 551– marker for impaired perceptual structural encoding 565. doi:10.1162/jocn.1996.8.6.551 Chan, A. W. Y., Kravitz, D. J., Truong, S., Arizpe, J., & when these signals are disrupted. It has previously Baker, C. I. (2010). Cortical representations of bodies been shown that relative to intact faces, the N170 to and faces are strongest in commonly experienced faces without eyes is also delayed (Eimer, 1998; Itier configurations. Nature Neuroscience, 13, 417–418. et al., 2007), analogous to the N170 delay elicited by doi:10.1038/nn.2502 contrast-inverting the eye region in the present study. De Lissa, P., McArthur, G., Hawelka, S., Palermo, R., fl Mahajan, Y., & Hutzler, F. (2014). Fixation location on This effect may re ect a similar disruption of face- upright and inverted faces modulates the N170. processing mechanisms that are selectively tuned to Neuropsychologia, 57,1–11. doi:10.1016/j. the eye region. Our additional finding that N170 neuropsychologia.2014.02.006 modulations triggered by contrast-inverting the eyes Duchaine, B., & Nakayama, K. (2006). The Cambridge face or the rest of the face are strongly affected by fixation memory test: Results for neurologically intact fl individuals and an investigation of its validity using location may re ect systematic retinotopic biases that inverted face stimuli and prosopagnosic participants. are linked to the special role of contrast-related signals Neuropsychologia, 44, 576–585. doi:10.1016/j. from the eye region. Because fixation near the eyes is neuropsychologia.2005.07.001 FACE PERCEPTION AND CONTRAST INVERSION 137

Eimer, M. (1998). Does the face-specific N170 component Vision Research, 39, 4003–4009. doi:10.1016/S0042- reflect the activity of a specialized eye processor? 6989(99)00109-1 Neuroreport, 9, 2945–2948. doi:10.1097/00001756- Liu, C. H., Collin, C. A., & Chaudhuri, A. (2000). Does 199809140-00005 face recognition rely on encoding of 3-D surface? Eimer, M. (2000). Event-related brain potentials distinguish Examining the role of shape-from-shading and shape- processing stages involved in face perception and from-stereo. Perception, 29, 729–743. doi:10.1068/ recognition. Clinical Neurophysiology, 111, 694–705. p3065 doi:10.1016/S1388-2457(99)00285-0 Nederhouser, M., Yue, X., Mangini, M. C., & Biederman, I. Eimer, M. (2011). The face-sensitive N170 component of (2007). The deleterious effect of contrast reversal on the event-related brain potential. In A. J. Calder, G. recognition is unique to faces, not objects. Vision Rhodes, M. H. Johnson, & J. V. Haxby (Eds.), The Research, 47, 2134–2142. doi:10.1016/j. oxford handbook of face perception (pp. 329–344). visres.2007.04.007 Oxford: Oxford University Press. Nemrodov, D., Anderson, T., Preston, F. F., & Itier, R. J. Galper, R. E. (1970). Recognition of faces in photographic (2014). Early sensitivity for eyes within faces: A new negative. Psychonomic Science, 19, 207–208. neuronal account of holistic and featural processing. doi:10.3758/BF03328777 NeuroImage, 97,81–94. doi:10.1016/j.neuroimage. Gandhi, T., Suresh, N., & Sinha, P. (2012). EEG responses 2014.04.042 to facial contrast-chimeras. Journal of Integrative Orban de Xivry, J.J., Ramon, M., Lefèvre, P., & Rossion, B. Neuroscience, 11, 201–211. doi:10.1142/ (2008). Reduced fixation on the upper area of personally S021963521250015X familiar faces following acquired . Gilad, S., Meng, M., & Sinha, P. (2009). Role of ordinal Journal of Neuropsychology, 2, 245–268. doi:10.1348/ contrast relationships in face encoding. Proceedings of 174866407X260199 the National Academy of Sciences, 106, 5353–5358. Peterson, M. F., & Eckstein, M. P. (2012). Looking just doi:10.1073/pnas.0812396106 below the eyes is optimal across face recognition tasks. Hsiao, J. H., & Cottrell, G. (2008). Two fixations suffice in Proceedings of the National Academy of Sciences, 109, face recognition. Psychological Science, 19, 998–1006. E3314–E3323. doi:10.1073/pnas.1214269109 doi:10.1111/j.1467-9280.2008.02191.x Rossion, B. (2009). Distinguishing the cause and Itier, R. J., Alain, C., Sedore, K., & McIntosh, A. R. (2007). consequence of face inversion: The perceptual field Early face processing specificity: It’s in the eyes! It’sin hypothesis. Acta Psychologica, 132, 300–312. the eyes! Journal of Cognitive Neuroscience, 19, 1815– doi:10.1016/j.actpsy.2009.08.002 1826. doi:10.1162/jocn.2007.19.11.1815 Rossion, B., Delvenne, J. F., Debatisse, D., Goffaux, V., Itier, R. J., Latinus, M., & Taylor, M. J. (2006). Face, eye Bruyer, R., Crommelinck, M., & Guérit, J. M. (1999). and object early processing: What is the face specificity? Spatio-temporal localization of the face inversion effect: NeuroImage, 29, 667–676. doi:10.1016/j. An event-related potentials study. Biological neuroimage.2005.07.041 Psychology, 50, 173–189. doi:10.1016/S0301-0511(99) Itier, R. J., & Taylor, M. J. (2002). Inversion and contrast 00013-7 polarity reversal affect both encoding and recognition Russell, R., Sinha, P., Biederman, I., & Nederhouser, M. processes of unfamiliar faces: A repetition study using (2006). Is pigmentation important for face recognition? ERPs. Neuroimage, 15, 353–372. doi:10.1006/ Evidence from contrast negation. Perception, 35, 749– nimg.2001.0982 759. doi:10.1068/p5490 Johnston, A., Hill, H., & Carman, N. (1992). Recognising Sinha, P. (2002). Qualitative representations for recognition. faces: Effects of lighting direction, inversion, and In H. H. Bulthoff, S.W. Lee, T. A. Poggio, & brightness reversal. Perception, 21, 365–375. C. Wallraven (Eds.), In biologically motivated Downloaded by [Radcliffe Infirmary] at 02:12 29 July 2016 doi:10.1068/p210365 computer vision (pp. 249–262). Berlin: Springer. Kemp, R., Pike, G., White, P., & Musselman, A. (1996). Sormaz,M.,Andrews,T.J.,&Young,A.W.(2013). Perception and recognition of normal and negative faces: Contrast negation and the importance of the eye The role of shape from shading and pigmentation cues. region for holistic representations of facial identity. Perception, 25,37–52. doi:10.1068/p250037 Journal of Experimental Psychology: Laguesse, R., Dormal, G., Biervoye, A., Kuefner, D., & Perception and Performance, 39, 1667–1677. Rossion, B. (2012). Extensive visual training in doi:10.1037/a0032449 adulthood significantly reduces the face inversion Vuong, Q. C., Peissig, J. J., Harrison, M. C., & Tarr, M. J. effect. Journal of Vision, 12, 14. doi:10.1167/12.10.14 (2005). The role of surface pigmentation for recognition Liu, C. H., Collin, C. A., Burton, A. M., & Chaudhuri, A. revealed by contrast reversal in faces and greebles. (1999). Lighting direction affects recognition of Vision Research, 45, 1213–1223. doi:10.1016/j. untextured faces in photographic positive and negative. visres.2004.11.015